<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Infra on ku5e | Cybersecurity Portfolio</title><link>https://ku5e.com/tags/infra/</link><description>Recent content in Infra on ku5e | Cybersecurity Portfolio</description><generator>Hugo -- 0.162.1</generator><language>en-us</language><lastBuildDate>Tue, 12 May 2026 11:14:00 +0000</lastBuildDate><atom:link href="https://ku5e.com/tags/infra/index.xml" rel="self" type="application/rss+xml"/><item><title>Frontier Models Are Overkill for Most Production Workloads</title><link>https://ku5e.com/blog/frontier-models-are-overkill-for-most-production-workloads/</link><pubDate>Tue, 12 May 2026 11:14:00 +0000</pubDate><guid>https://ku5e.com/blog/frontier-models-are-overkill-for-most-production-workloads/</guid><description>Open weight models are close enough to frontier performance for most production workloads. What changed this week, what the frontier models still win on, and what the practical question actually is.</description></item><item><title>Claude Code on DeepSeek: 17x Cheaper</title><link>https://ku5e.com/blog/claude-code-on-deepseek-17x-cheaper/</link><pubDate>Thu, 07 May 2026 10:27:00 +0000</pubDate><guid>https://ku5e.com/blog/claude-code-on-deepseek-17x-cheaper/</guid><description>DeepClaude routes Claude Code&amp;#39;s tool calls through DeepSeek instead of Anthropic&amp;#39;s models. Same tool ecosystem, approximately 17x lower cost. How it works and where it falls short.</description></item><item><title>DeepSeek V4 Broke the Pricing Argument</title><link>https://ku5e.com/blog/deepseek-v4-broke-the-pricing-argument/</link><pubDate>Sun, 03 May 2026 21:17:00 +0000</pubDate><guid>https://ku5e.com/blog/deepseek-v4-broke-the-pricing-argument/</guid><description>DeepSeek V4 is open weight, near-SOTA on benchmarks, and costs $1.74 per million input tokens. The pricing argument for closed frontier models just got harder to make.</description></item><item><title>I Built a Trading Bot That Runs Its LLM on a Jetson in My Closet</title><link>https://ku5e.com/blog/i-built-a-trading-bot-that-runs-its-llm-on-a-jetson-in-my-closet/</link><pubDate>Thu, 30 Apr 2026 21:03:00 +0000</pubDate><guid>https://ku5e.com/blog/i-built-a-trading-bot-that-runs-its-llm-on-a-jetson-in-my-closet/</guid><description>How I built a paper trading bot that runs its LLM on a Jetson Orin Nano in my home lab — architecture, the bug that would have been catastrophic, and why the model is explicitly blocked from touching execution.</description></item></channel></rss>